Examining brain microstructure using structure tensor analysis of histological sections. Neuroimage 2012 Oct 15;63(1):1-10
Date
07/05/2012Pubmed ID
22759994DOI
10.1016/j.neuroimage.2012.06.042Scopus ID
2-s2.0-84864039331 (requires institutional sign-in at Scopus site) 117 CitationsAbstract
The mammalian central nervous system has a tremendous structural complexity, and diffusion tensor imaging (DTI) is unique in its ability to extract microstructural tissue properties at a macroscopic scale. However, despite its widespread use and applications in clinical and research settings, accurate validation of DTI has notoriously lagged the advances in image acquisition and analysis. In this report, we demonstrate an approach to visualize and quantify the microscopic features of histological sections on multiple length scales using techniques derived from image texture analysis. Structure tensor (ST) analysis was applied to fluorescence microscopy images of rat brain sections to visualize and quantify tissue microstructure. Images were digitally color-coded based on the local orientation in the pixelwise ST implementation, which allowed direct visualization of white matter complexity at the microscopic level. A piecewise ST algorithm was also employed to quantify anisotropy and orientation at a resolution comparable to that typically acquired with DTI. Anisotropy measured with ST analysis of stained histological sections was highly correlated with anisotropy measured by ex vivo DTI of the same brains (R(2)=0.92). Furthermore, angular histograms, or Fiber Orientation Distributions (FODs), were computed to mimic similar measures derived from high angular resolution diffusion imaging methods. The FODs for each pixel were fit to a mixture of von Mises distributions to identify putative regions of multiple fiber populations (i.e. crossing fibers). Despite its current application to two-dimensional microscopy, the ST analysis is a novel approach to visualize and quantify microstructure in the central nervous system in both health and disease, and advances the available set of tools for validating DTI and other diffusion MRI techniques.
Author List
Budde MD, Frank JAAuthor
Matthew Budde PhD Associate Professor in the Neurosurgery department at Medical College of WisconsinMESH terms used to index this publication - Major topics in bold
AlgorithmsAnimals
Brain
Diffusion Magnetic Resonance Imaging
Female
Image Enhancement
Image Interpretation, Computer-Assisted
Pattern Recognition, Automated
Rats
Rats, Sprague-Dawley
Reproducibility of Results
Sensitivity and Specificity